MindSculpt enables users to generate a wide range of hybrid geometries in Grasshopper in real time simply by thinking about those geometries. This design tool combines a brain-computer interface (BCI) with the parametric design platform Grasshopper, creating an intuitive design workflow that shortens the latency between ideation and implementation compared to traditional computer-aided design tools based on mouse-and-keyboard paradigms. The project arises from transdisciplinary research between neuroscience and architecture, with the goal of building a cyber-human collaborative tool that is capable of leveraging the complex and fluid nature of thinking in the design process. MindSculpt applies a supervised machine-learning approach, based on the support vector machine model (SVM), to identify patterns of brain waves that occur in EEG data when participants mentally rotate four different solid geometries. The researchers tested MindSculpt with participants who had no prior experience in design and found that the tool was enjoyable to use and could contribute to design ideation and artistic endeavors.

翻译：MindSculpt使用户能够在Grasshopper中通过单纯想象几何形态，实时生成多种混合几何体。该设计工具将脑机接口与参数化设计平台Grasshopper相结合，创建了直观的设计工作流，相较于基于鼠标-键盘范式的传统计算机辅助设计工具，有效缩短了从构思到实现的延迟。该项目源于神经科学与建筑学的跨学科研究，旨在构建一种能够充分利用设计过程中思维的复杂性与流动性的网络-人类协作工具。MindSculpt采用基于支持向量机模型的监督式机器学习方法，识别参与者在心理旋转四种不同实体几何体时脑电图数据中出现的脑波模式。研究团队对无设计经验的参与者进行了MindSculpt测试，发现该工具既具使用趣味性，又能促进设计构思与艺术创作。